As shown in FIG. 1, multiple service gateway devices are deployed in a distributed way at enterprise exits, and traffic data from the Internet is shared among the multiple service gateway devices (e.g., service gateways A, B and C in FIG. 1) on an egress router by means of equivalent routing loads. Service transmission rates (also referred to as traffic) to destination hosts need to be limited on the service gateways due to the needs for resisting Distributed Denial of Service (DDoS) attacks or bandwidth limitation, to ensure that transmission rates of services entering an intranet do not exceed bandwidths purchased by users, thus avoiding an impact on intranet links.
However, as each service gateway limits flows (i.e., limits the service transmission rates) independently in such a distributed deployment scenario, how to ensure that the total traffic of the flows (i.e., a total service transmission rate for an identical destination address) that are limited in parallel satisfies an expected threshold after the flows pass through multiple service gateways needs to considered.
However, no effective solution has currently been proposed for the foregoing problem.